-
STUDY: Socially Aware Temporally Causal Decoder Recommender Systems
Authors:
Eltayeb Ahmed,
Diana Mincu,
Lauren Harrell,
Katherine Heller,
Subhrajit Roy
Abstract:
Recommender systems are widely used to help people find items that are tailored to their interests. These interests are often influenced by social networks, making it important to use social network information effectively in recommender systems. This is especially true for demographic groups with interests that differ from the majority. This paper introduces STUDY, a Socially-aware Temporally caU…
▽ More
Recommender systems are widely used to help people find items that are tailored to their interests. These interests are often influenced by social networks, making it important to use social network information effectively in recommender systems. This is especially true for demographic groups with interests that differ from the majority. This paper introduces STUDY, a Socially-aware Temporally caUsal Decoder recommender sYstem. STUDY introduces a new socially-aware recommender system architecture that is significantly more efficient to learn and train than existing methods. STUDY performs joint inference over socially connected groups in a single forward pass of a modified transformer decoder network. We demonstrate the benefits of STUDY in the recommendation of books for students who are dyslexic, or struggling readers. Dyslexic students often have difficulty engaging with reading material, making it critical to recommend books that are tailored to their interests. We worked with our non-profit partner Learning Ally to evaluate STUDY on a dataset of struggling readers. STUDY was able to generate recommendations that more accurately predicted student engagement, when compared with existing methods.
△ Less
Submitted 5 September, 2023; v1 submitted 2 June, 2023;
originally announced June 2023.
-
Predicting Cardiovascular Disease Risk using Photoplethysmography and Deep Learning
Authors:
Wei-Hung Weng,
Sebastien Baur,
Mayank Daswani,
Christina Chen,
Lauren Harrell,
Sujay Kakarmath,
Mariam Jabara,
Babak Behsaz,
Cory Y. McLean,
Yossi Matias,
Greg S. Corrado,
Shravya Shetty,
Shruthi Prabhakara,
Yun Liu,
Goodarz Danaei,
Diego Ardila
Abstract:
Cardiovascular diseases (CVDs) are responsible for a large proportion of premature deaths in low- and middle-income countries. Early CVD detection and intervention is critical in these populations, yet many existing CVD risk scores require a physical examination or lab measurements, which can be challenging in such health systems due to limited accessibility. Here we investigated the potential to…
▽ More
Cardiovascular diseases (CVDs) are responsible for a large proportion of premature deaths in low- and middle-income countries. Early CVD detection and intervention is critical in these populations, yet many existing CVD risk scores require a physical examination or lab measurements, which can be challenging in such health systems due to limited accessibility. Here we investigated the potential to use photoplethysmography (PPG), a sensing technology available on most smartphones that can potentially enable large-scale screening at low cost, for CVD risk prediction. We developed a deep learning PPG-based CVD risk score (DLS) to predict the probability of having major adverse cardiovascular events (MACE: non-fatal myocardial infarction, stroke, and cardiovascular death) within ten years, given only age, sex, smoking status and PPG as predictors. We compared the DLS with the office-based refit-WHO score, which adopts the shared predictors from WHO and Globorisk scores (age, sex, smoking status, height, weight and systolic blood pressure) but refitted on the UK Biobank (UKB) cohort. In UKB cohort, DLS's C-statistic (71.1%, 95% CI 69.9-72.4) was non-inferior to office-based refit-WHO score (70.9%, 95% CI 69.7-72.2; non-inferiority margin of 2.5%, p<0.01). The calibration of the DLS was satisfactory, with a 1.8% mean absolute calibration error. Adding DLS features to the office-based score increased the C-statistic by 1.0% (95% CI 0.6-1.4). DLS predicts ten-year MACE risk comparable with the office-based refit-WHO score. It provides a proof-of-concept and suggests the potential of a PPG-based approach strategies for community-based primary prevention in resource-limited regions.
△ Less
Submitted 9 May, 2023;
originally announced May 2023.
-
PADLL: Taming Metadata-intensive HPC Jobs Through Dynamic, Application-agnostic QoS Control
Authors:
Ricardo Macedo,
Mariana Miranda,
Yusuke Tanimura,
Jason Haga,
Amit Ruhela,
Stephen Lien Harrell,
Richard Todd Evans,
José Pereira,
João Paulo
Abstract:
Modern I/O applications that run on HPC infrastructures are increasingly becoming read and metadata intensive. However, having multiple concurrent applications submitting large amounts of metadata operations can easily saturate the shared parallel file system's metadata resources, leading to overall performance degradation and I/O unfairness. We present PADLL, an application and file system agnost…
▽ More
Modern I/O applications that run on HPC infrastructures are increasingly becoming read and metadata intensive. However, having multiple concurrent applications submitting large amounts of metadata operations can easily saturate the shared parallel file system's metadata resources, leading to overall performance degradation and I/O unfairness. We present PADLL, an application and file system agnostic storage middleware that enables QoS control of data and metadata workflows in HPC storage systems. It adopts ideas from Software-Defined Storage, building data plane stages that mediate and rate limit POSIX requests submitted to the shared file system, and a control plane that holistically coordinates how all I/O workflows are handled. We demonstrate its performance and feasibility under multiple QoS policies using synthetic benchmarks, real-world applications, and traces collected from a production file system. Results show that PADLL can enforce complex storage QoS policies over concurrent metadata-aggressive jobs, ensuring fairness and prioritization.
△ Less
Submitted 23 March, 2023; v1 submitted 13 February, 2023;
originally announced February 2023.
-
Improved optical standing-wave beam splitters for dilute Bose-Einstein condensates
Authors:
Mary Clare Cassidy,
Malcolm G. Boshier,
Lee E. Harrell
Abstract:
Bose-Einstein condensate (BEC)-based atom interferometry exploits low temperatures and long coherence lengths to facilitate high-precision measurements. Progress in atom interferometry promises improvements in navigational devices like gyroscopes and accelerometers, as well as applications in fundamental physics such as accurate determination of physical constants. Previous work demonstrates that…
▽ More
Bose-Einstein condensate (BEC)-based atom interferometry exploits low temperatures and long coherence lengths to facilitate high-precision measurements. Progress in atom interferometry promises improvements in navigational devices like gyroscopes and accelerometers, as well as applications in fundamental physics such as accurate determination of physical constants. Previous work demonstrates that beam splitters and mirrors for coherent manipulation of dilute BEC momentum in atom interferometers can be implemented with sequences of non-resonant standing-wave light pulses. While previous work focuses on the optimization of the optical pulses' amplitude and duration to produce high-order momentum states with high fidelity, we explore how varying the shape of the optical pulses affects optimal beam-splitter performance, as well as the effect of pulse shape on the sensitivity of optimized parameters in achieving high fidelity in high-momentum states. In simulations of two-pulse beam splitters utilizing optimized square, triangle, and sinc-squared pulse shapes applied to dilute BECs, we, in some cases, reduce parameter sensitivity by an order of magnitude while maintaining fidelity.
△ Less
Submitted 23 November, 2021;
originally announced November 2021.
-
VPIC 2.0: Next Generation Particle-in-Cell Simulations
Authors:
Robert Bird,
Nigel Tan,
Scott V. Luedtke,
Stephen Lien Harrell,
Michela Taufer,
Brian Albright
Abstract:
VPIC is a general purpose Particle-in-Cell simulation code for modeling plasma phenomena such as magnetic reconnection, fusion, solar weather, and laser-plasma interaction in three dimensions using large numbers of particles. VPIC's capacity in both fidelity and scale makes it particularly well-suited for plasma research on pre-exascale and exascale platforms. In this paper we demonstrate the uniq…
▽ More
VPIC is a general purpose Particle-in-Cell simulation code for modeling plasma phenomena such as magnetic reconnection, fusion, solar weather, and laser-plasma interaction in three dimensions using large numbers of particles. VPIC's capacity in both fidelity and scale makes it particularly well-suited for plasma research on pre-exascale and exascale platforms. In this paper we demonstrate the unique challenges involved in preparing the VPIC code for operation at exascale, outlining important optimizations to make VPIC efficient on accelerators. Specifically, we show the work undertaken in adapting VPIC to exploit the portability-enabling framework Kokkos and highlight the enhancements to VPIC's modeling capabilities to achieve performance at exascale. We assess the achieved performance-portability trade-off through a suite of studies on nine different varieties of modern pre-exascale hardware. Our performance-portability study includes weak-scaling runs on three of the top ten TOP500 supercomputers, as well as a comparison of low-level system performance of hardware from four different vendors.
△ Less
Submitted 25 February, 2021;
originally announced February 2021.
-
Lagrangian and impedance spectroscopy treatments of electric force microscopy
Authors:
Ryan P. Dwyer,
Lee E. Harrell,
John A. Marohn
Abstract:
Scanning probe microscopy is often extended beyond topographic imaging to study electrical forces and sample properties, with the most widely used experiment being frequency-modulated Kelvin probe force microscopy. The equations commonly used to interpret this experiment, however, rely on two hidden assumptions: (1) the tip charge oscillates in phase with the cantilever motion to keep the tip volt…
▽ More
Scanning probe microscopy is often extended beyond topographic imaging to study electrical forces and sample properties, with the most widely used experiment being frequency-modulated Kelvin probe force microscopy. The equations commonly used to interpret this experiment, however, rely on two hidden assumptions: (1) the tip charge oscillates in phase with the cantilever motion to keep the tip voltage constant, and (2) any changes in the tip-sample interaction happen slowly. Starting from an electro-mechanical model of the cantilever-sample interaction, we use Lagrangian mechanics to derive coupled equations of motion for the cantilever position and charge. This general approach rigorously describes scanned probe experiments even in the case when the usual assumptions of fast tip charging and slowly changing samples properties are violated. We develop a Magnus-expansion approximation to illustrate how abrupt changes in the tip-sample interaction cause abrupt changes in the cantilever amplitude and phase. We show that feedback-free time-resolved electric force microscopy cannot uniquely determine sub-cycle photocapacitance dynamics. We then use first-order perturbation theory to relate cantilever frequency shift and dissipation to the sample impedance even when the tip charge oscillates out of phase with the cantilever motion. Analogous to the treatment of impedance spectroscopy in electrochemistry, we apply this approximation to determine the cantilever frequency shift and dissipation for an arbitrary sample impedance in (broadband) local dielectric spectroscopy experiments. The general approaches we develop provide a path forward for rigorously modeling the coupled motion of the cantilever position and charge in the wide range of electrical scanned probe microscopy experiments where the hidden assumptions of the conventional equations are violated or inapplicable.
△ Less
Submitted 10 February, 2019; v1 submitted 3 July, 2018;
originally announced July 2018.
-
Wavefunction collapse through backaction of counting weakly interacting photons
Authors:
Lee E. Harrell
Abstract:
We apply the formalism of quantum measurement theory to the idealized measurement of the position of a particle with an optical interferometer, finding that the backaction of counting entangled photons systematically collapses the particle's wavefunction toward a narrow Gaussian wavepacket at the location $x_\mathrm{est}$ determined by the measurement without appeal to environmental decoherence or…
▽ More
We apply the formalism of quantum measurement theory to the idealized measurement of the position of a particle with an optical interferometer, finding that the backaction of counting entangled photons systematically collapses the particle's wavefunction toward a narrow Gaussian wavepacket at the location $x_\mathrm{est}$ determined by the measurement without appeal to environmental decoherence or other spontaneous collapse mechanism. Further, the variance in the particle's position, as calculated from the post-measurement wavefunction agrees precisely with shot-noise limited uncertainty of the measured $x_\mathrm{est}$. Both the identification of the absolute square of the particle's initial wavefunction as the probability density for $x_\mathrm{est}$ and the de Broglie hypothesis emerge as consequences of interpreting the intensity of the optical field as proportional to the probability of detecting a photon. Linear momentum information that is encoded in the particle's initial wavefunction survives the measurement, and the pre-measurement expectation values are preserved in the ensemble average.
△ Less
Submitted 4 April, 2016; v1 submitted 8 January, 2016;
originally announced January 2016.
-
Practical recipes for the model order reduction, dynamical simulation, and compressive sampling of large-scale open quantum systems
Authors:
John A. Sidles,
Joseph L. Garbini,
Lee E. Harrell,
Alfred O. Hero,
Jonathan P. Jacky,
Joseph R. Malcomb,
Anthony G. Norman,
Austin M. Williamson
Abstract:
This article presents numerical recipes for simulating high-temperature and non-equilibrium quantum spin systems that are continuously measured and controlled. The notion of a spin system is broadly conceived, in order to encompass macroscopic test masses as the limiting case of large-j spins. The simulation technique has three stages: first the deliberate introduction of noise into the simulati…
▽ More
This article presents numerical recipes for simulating high-temperature and non-equilibrium quantum spin systems that are continuously measured and controlled. The notion of a spin system is broadly conceived, in order to encompass macroscopic test masses as the limiting case of large-j spins. The simulation technique has three stages: first the deliberate introduction of noise into the simulation, then the conversion of that noise into an equivalent continuous measurement and control process, and finally, projection of the trajectory onto a state-space manifold having reduced dimensionality and possessing a Kahler potential of multi-linear form. The resulting simulation formalism is used to construct a positive P-representation for the thermal density matrix. Single-spin detection by magnetic resonance force microscopy (MRFM) is simulated, and the data statistics are shown to be those of a random telegraph signal with additive white noise. Larger-scale spin-dust models are simulated, having no spatial symmetry and no spatial ordering; the high-fidelity projection of numerically computed quantum trajectories onto low-dimensionality Kahler state-space manifolds is demonstrated. The reconstruction of quantum trajectories from sparse random projections is demonstrated, the onset of Donoho-Stodden breakdown at the Candes-Tao sparsity limit is observed, a deterministic construction for sampling matrices is given, and methods for quantum state optimization by Dantzig selection are given.
△ Less
Submitted 13 May, 2008;
originally announced May 2008.
-
170 Nanometer Nuclear Magnetic Resonance Imaging using Magnetic Resonance Force Microscopy
Authors:
Kent R. Thurber,
Lee E. Harrell,
Doran D. Smith
Abstract:
We demonstrate one-dimensional nuclear magnetic resonance imaging of the semiconductor GaAs with 170 nanometer slice separation and resolve two regions of reduced nuclear spin polarization density separated by only 500 nanometers. This is achieved by force detection of the magnetic resonance, Magnetic Resonance Force Microscopy (MRFM), in combination with optical pum** to increase the nuclear…
▽ More
We demonstrate one-dimensional nuclear magnetic resonance imaging of the semiconductor GaAs with 170 nanometer slice separation and resolve two regions of reduced nuclear spin polarization density separated by only 500 nanometers. This is achieved by force detection of the magnetic resonance, Magnetic Resonance Force Microscopy (MRFM), in combination with optical pum** to increase the nuclear spin polarization. Optical pum** of the GaAs creates spin polarization up to 12 times larger than the thermal nuclear spin polarization at 5 K and 4 T. The experiment is sensitive to sample volumes containing $\sim 4 \times 10^{11}$ $^{71}$Ga$/\sqrt{Hz}$. These results demonstrate the ability of force-detected magnetic resonance to apply magnetic resonance imaging to semiconductor devices and other nanostructures.
△ Less
Submitted 23 October, 2002;
originally announced October 2002.
-
Temperature measurement at the end of a cantilever using oxygen paramagnetism in solid air
Authors:
Kent R. Thurber,
Lee E. Harrell,
Doran D. Smith
Abstract:
We demonstrate temperature measurement of a sample attached to the end of a cantilever using cantilever magnetometry of solid air ``contamination'' of the sample surface. In experiments like our Magnetic Resonance Force Microscopy (MRFM), the sample is mounted at the end of a thin cantilever with small thermal conductance. Thus, the sample can be at a significantly different temperature than the…
▽ More
We demonstrate temperature measurement of a sample attached to the end of a cantilever using cantilever magnetometry of solid air ``contamination'' of the sample surface. In experiments like our Magnetic Resonance Force Microscopy (MRFM), the sample is mounted at the end of a thin cantilever with small thermal conductance. Thus, the sample can be at a significantly different temperature than the bulk of the instrument. Using cantilever magnetometry of the oxygen paramagnetism in solid air provides the temperature of the sample, without any modifications to our MRFM (Magnetic Resonance Force Microscopy) apparatus.
△ Less
Submitted 23 October, 2002;
originally announced October 2002.
-
Spin polarization contrast observed in GaAs by force-detected nuclear magnetic resonance
Authors:
Kent R. Thurber,
Lee E. Harrell,
Raul Fainchtein,
Doran D. Smith
Abstract:
We applied the recently developed technique of force-detected nuclear magnetic resonance (NMR) to observe 71Ga, 69Ga, and 75As in GaAs. The nuclear spin-lattice relaxation time is 21$\pm$5 min for 69Ga at $\sim 5$ K and 4.6 Tesla. We have exploited this long relaxation time to first create and then observe spatially varying nuclear spin polarization within the sample, demonstrating a new form of…
▽ More
We applied the recently developed technique of force-detected nuclear magnetic resonance (NMR) to observe 71Ga, 69Ga, and 75As in GaAs. The nuclear spin-lattice relaxation time is 21$\pm$5 min for 69Ga at $\sim 5$ K and 4.6 Tesla. We have exploited this long relaxation time to first create and then observe spatially varying nuclear spin polarization within the sample, demonstrating a new form of contrast for magnetic resonance force microscopy (MRFM). Such nuclear spin contrast could be used to indirectly image electron spin polarization in GaAs-based spintronic devices.
△ Less
Submitted 14 January, 2002;
originally announced January 2002.
